The overarching objective of the proposed research is to evaluate whether changes in drinking identity (how much one associates one's self with drinking) can reduce hazardous drinking (HD; heavy alcohol use and negative alcohol-related consequences) among current and graduating college students. This work addresses two key priorities of NIAAA's strategic plan: (1) to identify factors that enable some young adults to transition out of HD without formal treatment, and (2) to test whether these factors can be manipulated to aid in the development of novel behavioral approaches to reduce HD. The public health burden of young adult HD is substantial: despite decades of research, HD is largely unabated and appears to be increasing in severity among college populations. While progress has been made towards identifying HD risk factors and developing interventions, HD levels remain persistently high. Moreover, recent findings question the effectiveness of interventions aimed at reducing college student HD, highlighting the need for improvement. We propose that drinking identity is a promising factor to investigate as a potential cognitive mechanism that contributes to the natural reductions in HD that occur for many college graduates. Recent findings have revealed drinking identity to be a robust predictor of college student HD cross-sectionally and over time, and changes in drinking identity lead to changes in alcohol consumption and vice-versa. Further, cognitions related to identity, drinking, and substance use can be manipulated, which can lead to increased self-efficacy to refuse a substance and reduced craving for that substance. Despite this promise, drinking identity has not been evaluated as a predictor of the transition out of HD; no studies have sought to change drinking identity directly; and no college HD interventions directly target drinking identity. The proposed studies, therefore, seek to bridge this gap by evaluating naturalistic (Study 1) and experimentally-induced (Study 2) change in drinking identity. The naturalistic study will follow a sample of 400 soon-to-graduate college HDs and evaluate their drinking identity and drinking behaviors at baseline and every 4 months for the 2-year period after graduation. This will allow for examination of within- and between-person change in drinking identity and HD at a key developmental period when reductions in HD occur. The experimental study will manipulate drinking identity (via narrative writing tasks) to increase self-efficacy and decrease alcohol craving and ultimately, reduce HD. A sample of 328 college student HDs will be recruited for a 3-session, lab-based study that includes 2-week, 1-month, and 3- month follow ups to test the potential for more durable, lasting change in drinking, self-efficacy, craving, and real-world drinking. The proposed research is particularly important because it evaluates a promising cognitive factor that is not currently targeted in existing behavioral treatments for HD, provides direct test of proposed mechanisms of change, and could ultimately lead to novel, brief interventions for reducing college student HD.